Device for combining and detecting shoulders and openings of glass bottles

Abstract

The utility model relates to a glass bottle shoulder and bottle opening combination detection device which comprises a detection table used for placing a to-be-detected glass bottle; the visual detection assembly comprises at least one industrial camera I and at least one industrial camera II, a matched lens I is arranged on the industrial camera I, a matched lens II is arranged on the industrial camera II, and the lens II is arranged at the position capable of effectively shooting the bottleneck area of the glass bottle; the backlight light source assembly is arranged on the periphery of the glass bottle detection position; the visual detection assembly, the backlight power supply and the visual processing module are all connected with the control module, and the visual processing module is used for processing and analyzing images shot by the visual detection assembly. According to the utility model, through a camera synchronous adjustment mechanism, camera configuration can be rapidly switched through preset parameters when products are changed, so that a large amount of adjustment time and workload are reduced, and the space utilization rate of a production workshop is improved.

Description

TECHNICAL FIELD

[0001] The utility model relates to glass bottle production detection technical field especially relates to glass bottle shoulder bottle opening combined detection device. BACKGROUND

[0002] On the production line of cosmetic glass bottles, product quality control faces many challenges. The traditional manual detection method has significant defects.

[0003] In the detection link, since the bottle opening and shoulder detection have similarity in operation process, camera adjustment is repeated every time the production is changed, which not only consumes a lot of time and energy, but also increases the wear and maintenance cost of hardware equipment, reduces the production efficiency, and becomes the key problem to be solved in the industry. UTILITY MODEL CONTENT

[0004] The utility model provides glass bottle shoulder bottle opening combined detection device in view of the deficiency of prior art, and the specific technical scheme is as follows:

[0005] Including detection table, for placing the glass bottle to be detected;

[0006] Visual detection component, including at least one industrial camera one and at least one industrial camera two, the industrial camera one is equipped with the lens one of being adapted, the industrial camera two is equipped with the lens two of being adapted, the lens one is arranged in the position that can effectively shoot glass bottle shoulder area, the lens two is arranged in the position that can effectively shoot glass bottle opening area;

[0007] Backlight light source component, it is configured at the periphery of glass bottle detection position, for providing the illumination light that penetrates glass bottle body, to enhance the contrast between glass bottle body and defect;

[0008] Visual processing module and control module, visual detection component, backlight power supply and visual processing module are all connected with control module, and the visual processing module is used for the image of visual detection component and carries out processing analysis.

[0009] Preferably, the backlight light source component is a uniform distribution of surrounding structure, and is surrounded around the periphery of the glass bottle detection position, and the backlight light source component is an LED lamp strip or a ring-shaped LED light source.

[0010] Preferably, it further includes feeding conveying component, the detection table is arranged at the end of feeding conveying component, and the feeding conveying component is used to convey glass bottle to the detection table, and the feeding conveying component is a conveyor and is connected with the control module.

[0011] Preferably, the proximity sensor is further arranged on the detection table, and the proximity sensor is connected with the visual detection assembly and the backlight light source assembly, and when the glass bottle reaches the detection table position, the proximity sensor can trigger the visual detection assembly and the backlight light source to operate simultaneously, and transmit a trigger signal to the control module to operate the visual processing module to detect.

[0012] Preferably, the first output device is further connected with the control module, and the first output device comprises a first electric telescopic rod arranged on one side of the detection table, and a collecting box is arranged on the side of the detection table away from the first electric telescopic rod, and the first electric telescopic rod can push the glass bottle into the collecting box.

[0013] Preferably, the second output device is further connected with the control module, and the second output device comprises a second electric telescopic rod arranged above the detection table.

[0014] Preferably, the display is further connected with the control module.

[0015] The beneficial effects of the present application are as follows:

[0016] The bottle shoulder bottle mouth of the present application is adjusted synchronously through a camera, and when the product is changed, the camera configuration is quickly switched through preset parameters, a large amount of adjustment time and workload is reduced, the bottle mouth bottle shoulder detection equipment is highly integrated, for example, two sets of independent detection equipment and supporting facilities originally needed to be purchased are now only one set of the device of the present application, the equipment floor area is reduced, the production workshop space utilization is improved, and a more efficient and economical production environment is created for enterprises. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The figure is a structural schematic diagram of the whole device.

[0018] Mark: 100, detection table; 1, industrial camera one; 11, lens one; 2, industrial camera two; 21, lens two; 3, backlight light source assembly; 4, feeding conveying assembly; 5, first electric telescopic rod; 6, collecting box; 7, second electric telescopic rod. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below with reference to the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not used to limit the present application.

[0020] Embodiment

[0021] Please refer to Figure 1, glass bottle shoulder bottle opening combined detection device, at least contains two groups of high-resolution industrial camera one and the matching lens. Industrial camera one 1 and lens are installed at a specific angle and position, the field of view accurately covers the glass bottle shoulder area, can capture the fine texture and potential defects of the part; industrial camera two 2 and lens two 21 focus on the bottle opening part, ensure the clear imaging of the complex structure of the bottle opening, provide high-quality image data for subsequent accurate detection.

[0022] The backlight source assembly 3 is designed as a surround backlight source, which is uniformly distributed around the glass bottle detection position. The stable high-intensity light emitted by the backlight source assembly 3 can penetrate the glass bottle body, enhance the contrast between the bottle body and the defects, effectively highlight the defect profile such as crack, gap, etc., assist the camera to obtain clearer image, and improve the detection accuracy.

[0023] Control module, vision processing module and display assembly: the control module is a high-performance computer terminal, the vision processing module is installed in the computer terminal, and the computer terminal is connected with the light source assembly through high-speed data transmission line. The vision processing module has built-in existing deep learning algorithm model, which has powerful image analysis processing capability; the display assembly directly shows the detection process data, image and final result in real time, which is convenient for the operator to monitor and manage the detection process.

[0024] In the vision processing module, the existing TensorFlow or PyTorch framework can be developed, which provides rich tools and function library, which is convenient for building complex neural network model for image recognition and segmentation task. Through the training of a large number of glass bottle defect samples, the software can learn the feature mode of various defects (such as crack, gap, etc.) of bottle shoulder and bottle opening. In actual detection, when the image captured by the camera is transmitted into the software, the model will analyze the image, identify the area where the defect may exist, and further extract the detailed information of the defect such as length, width, contrast, shape complexity, etc. through the subsequent existing Blob analysis algorithm. The device has significant innovation and competitive advantage.

[0025] Further, in the actual working process, the glass bottle is orderly conveyed to the detection table 100 by the feeding conveying assembly 4. When the glass bottle reaches the preset detection position, the proximity sensor is triggered, and the proximity sensor sends a signal to the control module. The control module synchronously instructs the industrial camera one 1 and two to start the lens one 11 and two to shoot, and simultaneously starts the backlight source assembly 3, and the camera instantly captures the image of the glass bottle shoulder and bottle opening, which is transmitted to the computer terminal at high speed through the data transmission line.

[0026] The visual processing module in the computer uses a pixel segmentation model based on deep learning to perform in-depth analysis of the images. This model, trained on a large number of glass bottle defect samples, can accurately identify various defect characteristics. During the analysis process, the image is first pre-processed, with operations such as noise reduction and contrast enhancement to improve image quality; then, image features are extracted using existing technologies such as convolutional neural networks, and compared with the defect feature library in the pre-trained model to locate potential defect areas. After the initial detection, the Blob analysis algorithm is used to further process the defect area, accurately calculating detailed parameter information such as the length, width, contrast, and shape complexity of the defect.

[0027] Based on the above analysis results, the software strictly judges the defect information according to the preset judgment rules. For example, for crack defects, set the length threshold and depth threshold, when the crack length exceeds the specified value or the depth reaches a certain level, the product is determined as NG (not qualified);

[0028] For notch defects, the area and edge irregularity are used as indicators for judgment. Finally, the control system drives the first output device to transport the NG products to the designated collection box 6, ensuring that qualified products smoothly enter the subsequent production link, and realizing an efficient and automated detection process.

[0029] Further, during installation in the production workshop layout, the detection device is positioned reasonably based on the direction and speed of the glass bottle conveyor on the production line. The feeding conveyor assembly 4 closely connects with the upstream production equipment, ensuring smooth transition and accurate positioning of the glass bottle in the detection area. After installation, the system is debugged. First, use standard defect sample glass bottles to calibrate the camera, adjust the focal length, aperture, and other parameters to make the bottle shoulder and mouth image clear and sharp; at the same time, test and optimize the sensor sensitivity to ensure accurate triggering of the detection program.

[0030] For different types of cosmetic glass bottles (such as cream bottles, perfume bottles, etc.), collect a large number of sample image data and use deep learning algorithms to train the model. During the training process, continuously adjust the model parameters to optimize the defect recognition ability, and after at least 1000 rounds of training and verification, the model can achieve ideal precision and stability in actual detection.

[0031] During daily production operations, the equipment continuously monitors the status of the feeding conveyor component 4. Once a glass bottle enters the detection area and triggers the proximity sensor, the detection process is immediately initiated. Images captured by an industrial camera are transmitted in real-time to a computer, where vision processing software rapidly analyzes and processes them. The detection results are displayed on the monitor screen within 0.5 seconds. For defective products, the first output device responds quickly, pushing them to the collection box 6 via an electric telescopic rod. Qualified products, driven by the second conveyor belt, continue flowing through the detection station 100 to the next connected conveyor for subsequent packaging processes. The entire process is fully automated and seamless, ensuring efficient, stable, and orderly production, effectively improving the quality and efficiency of the company's glass bottle production.

[0032] The above description is only a preferred embodiment of the present utility model and is not intended to limit the present utility model. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present utility model should be included within the protection scope of the present utility model.

Claims

1. A device for detecting the combined shoulder and mouth of a glass bottle, characterized in that, include: The testing station is used to place the glass bottles to be tested. A vision inspection component includes at least one industrial camera 1 and at least one industrial camera 2. The industrial camera 1 is equipped with a matching lens 1, and the industrial camera 2 is equipped with a matching lens 2. The lens 1 is positioned to effectively capture the shoulder area of ​​the glass bottle, and the lens 2 is positioned to effectively capture the mouth area of ​​the glass bottle. A backlight source assembly, which is configured around the glass bottle detection position, is used to provide illumination light that penetrates the glass bottle body to enhance the contrast between the glass bottle body and the defects. The system includes a vision processing module and a control module. The vision detection component, backlight power supply, and vision processing module are all connected to the control module. The vision processing module is used to process and analyze the images captured by the vision detection component.

2. The glass bottle shoulder and mouth merging detection device according to claim 1, characterized in that: The backlight source assembly is a uniformly distributed, surrounding structure that surrounds the detection position of the glass bottle. The backlight source assembly is an LED strip or a ring-shaped LED light source.

3. The glass bottle shoulder and mouth merging detection device according to claim 2, characterized in that: It also includes a feeding conveyor assembly, with the testing station located at the end of the feeding conveyor assembly. The feeding conveyor assembly is used to transport glass bottles to the testing station and is a conveyor connected to the control module.

4. The glass bottle shoulder and mouth merging detection device according to claim 3, characterized in that: It also includes a proximity sensor installed on the inspection platform. The proximity sensor is connected to the vision inspection component and the backlight source component. When the glass bottle reaches the position of the inspection platform, the proximity sensor can trigger the vision inspection component and the backlight source to operate simultaneously and transmit the trigger signal to the control module to run the vision processing module for inspection.

5. The glass bottle shoulder and mouth merging detection device according to claim 4, characterized in that: It also includes a first output device, which is connected to the control module. The first output device includes an electric telescopic rod located on one side of the testing platform. A collection box is provided on the side of the testing platform away from the electric telescopic rod. The electric telescopic rod can push the glass bottle into the collection box.

6. The glass bottle shoulder and mouth merging detection device according to claim 5, characterized in that: It also includes a second output device connected to the control module, the second output device including an electric telescopic rod located above the testing platform.

7. The glass bottle shoulder and mouth merging detection device according to claim 1, characterized in that: It also includes a display, which is connected to the control module.

Images